Research On Channel Estimations Of OFDM Communication System In PLC

Power Line Communication Technique is shortly called PLC, which has a promising future. It is a multi-media signal transmission method which makes use of the low-voltage power line to transfer the signal such as data, voice, picture and video. Because of its notable features of connecting urban and rural areas, no wiring and convenient connections and so on, PLC has being widely considered as one of the most competitive techniques to provide "last mile" solutions. As people's life becomes better, the need of the network speed is increasing. So how to improve the performance of the PLC communications system gets to be a research hotspot.Channel estimation is one of the key problems in the OFDM system. This thesis has improved two Channel estimation algorithms, that is, DFT and LMS. Improved DFT algorithm applies a decision threshold only within the length of the CP, based on which the CIR's threshold is set and the noise is removed further, therefore, the accuracy of channel estimation is improved. Selecting step size is pivotal for the LMS adaptive filter in the channel estimation. In the foundation of studying basic LMS algorithm and carrying on an analysis to LMS algorithm in the current power line channel, some problems about LMS algorithm are found in choosing parameter currently.The problem of parameter selection in the current power-line LMS channel estimation is adjusted in this paper, and momentum term of ANN (Artificial Neural Network) is first used to accelerate the convergence rate of LMS algorithm which makes it track the change of channel better. Simulation results show that the improved algorithm is superior to the original algorithms and is promising for practical applications.A simple OFDM system that transmitting static colorful pictures using MATALB is built finally in the thesis, which testify the validity of the proposed improved algorithm.